Computational Model of Functional Connectivity Distance Predicts Neural Alterations

Modeling brain signals play a crucial role in analyzing the brain's architecture, functions and associated disorders. This article aims to model the brain topology by exploring the relationship between complex neural correlates and acrlong FC-based distances. A computational model inspired by multivariate visibility graphs (VGs) algorithm and acrlong ED is proposed to analyze quantitatively the brain network data. When applied to resting-state acrlong EEG signals from three groups [typically developing (acrshort TD), autism spectrum disorder autism spectrum disorder (ASD), and epilepsy (E)], the network topological properties (e.g., global efficiency, modularity, small worldness, and betweenness centrality) demonstrate variations in connectivity distance probabilities among brain regions (e.g., frontal, temporal, parietal, and occipital) via the model's delay and connection distance parameters. The results showed a higher delay and skewed distribution toward short functional connections in ASD than in acrshort TD, while a lower delay in E than in ASD and acrshort TD. Additionally, ASD had more short-distance connections, while E had more long-distance connections compared to acrshort TD. ASD and E significantly overlapped over short-distance connections within the temporal lobe. In summary, the proposed model illustrates that delay parameter and connection distance obtained from brain network data have the potential to objectively identify and associate co-occurring neurological conditions (e.g., ASD and E).